Pharmacogenetics involves the study of host susceptibility and environmental exposures in cancer development. The unique contribution of this discipline is that common genes are studied, and the role of the environment is more explicit because the mechanism of the putative polymorphic gene of interest is known, and the study base is the population rather than the family. We use this as a starting point to design molecular epidemiology studies on lung, prostate, oral, gastric, and lymphoproliferative cancers and around special exposures such as dioxin. The major aim is to identify and better understand genetic components that contribute to major cancers. The studies involve interdisciplinary efforts and emphasize the integration of epidemiological, genetic, and laboratory approaches. Lung Cancer GEB has been a pioneer in studies aimed at identifying a genetic component to common cancers using epidemiological study designs. Early work emphasized polymorphic genes (e.g., CYP2D6, GSTM1) that are hypothesized to activate or detoxify carcinogens and thereby alter an individuals susceptibility to cancer when exposed to a specific agent. Lung cancer is a logical focus for these studies, because there is clear evidence of an environmental etiology (i.e., tobacco smoking), although only a minority of heavy smokers develop the disease. Furthermore, somatic gene mutations are well documented in lung cancer, and relatives of people who develop the cancer are also at increased risk. To better understand the postulated genetic component to smoking-related cancer we conduct case-control studies of lung cancer with extensive biospecimen collection; studies of determinants of the cancer in nonsmokers; studies of the cancer in different geographic, occupational, and ethnic groups; and studies of consecutive surgical cases with emphasis on tissue collection in order to characterize somatic genes. Families with lung and other smoking-related cancers have been studied, although suitable large, multigeneration families are difficult to accrue because of the death toll for smoking- related diseases. We participate in NCIs Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Study, which is a cohort study of 70,000 individuals in which large numbers of lung cancer cases will be identified over the next few years. Currently a study of the genetic determinants of chronic obstructive pulmonary disease is planned. In related research, we addressed the hypothesis of a genetic component to smoking itself. We have examined genetic polymorphisms affecting dopamine regulation andreceptor stimulation as candidates for influencing genetic susceptibility to cigarette smoking. Preliminary findings suggest that the dopamine transporter gene (DAT1), which governs the reuptake of dopamine from the neuronal synapse, and the D2 dopamine receptor gene (DRD2), which is a postsynaptic receptor, may interact to influence smoking behavior. This finding suggests that smoking may be influenced by an interplay among multiple genes affecting dopaminergic reuptake and receptor stimulation. Understanding the mechanisms of dopaminergic genes in smoking may facilitate development of improved strategies to prevent smoking and help smokers quit the habit. Integrated Studies In order to study genetic determinants of lung cancer and smoking we have designed a large new case-control study with sufficient power to detect gene-environment interactions, biospecimen collection to allow interdisciplinary study of biomarkers from normal and neoplastic tissue, tissue collection from lung cancer sib pairs, and the planned study of genetic determinants of smoking in controls. Other Tumors We are collaborate in studies of genetic susceptibility for a wide variety of other tumors. We have recently described a role for GSTT1 in prostate cancer and further work is underway to characterize NAT2 and other putative susceptibility genes in this malignancy. In a study of oral cancer, we found that an allele encoded by alcohol dehydrogenase, ADH3-1, oxidizes alcohol to its carcinogenic metabolite, acetaldehyde, more rapidly than enzymes encoded by the ADH3-2 allele. In a case-control study in a geographic area with elevated rates of oral cancer, we found that the overall risk for the disease was lower in individuals with the ADH3;1-2 or the ADH)3 2-2 genotypes than for those with the ADH3 1-1 genotype, particularly among the heaviest drinkers. These findings suggest that the ADH-3 locus influences drinkers risk of oral cancer, and implicate acetaldehyde in oral carcinogenesis. Special Populations Studies of unusual population exposures provide an opportunity to test for the role of mechanistically plausible genes in cancer causation. One example is the industrial accident that contaminated Seveso, Italy, with the highly toxic compound 2,3,7,8- tetrachloro-dibenzo(p)-dioxin (TCDD). Two decades after the incident, a population-based study was designed to determine current TCDD levels in healthy individuals from the contaminated zones and surrounding areas. We measured plasma TCDD and other polychlorinated compounds in a population-based sample of 62 individuals from the highly contaminated area and 59 individuals from a noncontaminated surrounding area. The study showed that TCDD levels were significantly higher in women than in men in both the contaminated and surrounding areas. The higher levels in women may have been partially due to increased body fat resulting in prolonged retention of the highly lipophilic compound. However, gender retained an independent association with TCDD, even after taking into account environmental factors related to TCDD levels, including smoking, distance from the accident site, and consumption of animals raised in the area at the time of the accident. Future research will examine the relation between risk of dioxin-related chloracne and cancer to dose as well as to expression of polymorphisms of candidate genes, including CYPlA1 and the ah receptor. Methodological Studies Our pharmacogenetics studies emphasize genotype-phenotype relationships, and extensive studies have been conducted to better understand the utility of probe drug assays for identifying gene polymorphisms. These studies have included ones on caffeine (for CYP1A2 and NAT2) and debrisoquine/ dextromethorphan (for CYP2D6) in various settings. In some cases, particular phenotyping approaches are compared; e.g., the caffeine breath test and simple consumption of caffeine and urine assay. Other studies have examined the disposition of nicotine in relation to the CYP2D6 and CYP2A6 genotypes. Research is being conducted on methodologic issues in biomarker studies in population studies, on practical and ethical aspects of biospecimen use in field studies, on optimum study design to detect gene-environment interaction, on the degree to which population stratification threatens studies of genetic susceptibility in epidemiologic studies, and on new DNA collection methodologies appropriate for large field studies.